Correlate structurally similar compounds with their known potentially cancer related endpoint using physico-chemical parameters such as partition coefficients, steric constants, or electronic constants. Correlate structurally diverse compounds with their known potentially cancer related endpoint using parameters such as substructure, interatomic distances, or molecular orbital calculations. In each area, determine the suitable physico-chemical parameters and mathematical techniques to correlate potentially cancer related endpoint data generated by NCI studies and in the open scientific literature. Accordingly, obtain physico-chemical parameters that are not available in the literature. From these correlations, gain insight into the mechanisms of carcinogenesis and increase the predicitive capability of anticipating the effects of untested compounds in biological systems.